Photolithography systems are known in the art that direct light beams onto a photosensitive surface covered by a mask, etching a desired pattern on the substrate corresponding to the void areas of the mask. Maskless photolithography systems are also known in the art as described in Singh-Gasson, Sangeet et al., Nature Biotechnology 17, 974–78, 1999. The system described in this article uses an off-axis light source coupled with a digital micromirror array to fabricate DNA chips containing probes for genes or other solid phase combinatorial chemistry to be performed in high-density microarrays.
A number of patents also exist which relate to maskless photolithography systems, including U.S. Pat. Nos. 5,870,176; 6,060,224; 6,177,980; and 6,251,550; all of which are incorporated herein by reference. While the previously described maskless photolithography systems address several of the problems associated with mask based photolithography systems, such as distortion and uniformity of images, problems still arise. Notably, in environments requiring rapid prototyping and limited production quantities, the advantages of maskless systems as a result of efficiencies derived from quantities of scale are not realized. Further, prior art references lack the ability to provide rapid prototyping.
In particular, alignment of patterns with respect to target substrates in maskless systems can be problematic. Various solutions have been proposed to mitigate the effect of alignment problems, including the digital shifting of the projected mask pattern to compensate for misalignment of a substrate. However, this technique requires that the substrate be closely aligned initially and is better suited for high volume production runs which incorporate automatic initial alignment systems. In a rapid prototyping, limited quantity environment, automated means of initial alignment are not cost effective.
In addition, conventional maskless alignment systems are normally limited to coplanar, two-dimensional alignment. However, there is a need in the art to create three-dimensional patterns on substrates. Creating three-dimensional patterns requires further alignment of the substrates in a third dimension perpendicular to the two coplanar dimensions. In the third dimension, computerized shifting of the mask pattern cannot compensate for misalignments in a direction parallel with an incident light beam. As a result, an ability to align a substrate in a third dimension in a rapid prototyping, reconfigurable environment is needed.
Another problem with maskless photolithography systems is that the mask pattern image projected is formed by pixels, instead of continuous lines. As a result, gaps may exist between adjacent pixels, which, when projected on a substrate, may allow the area between the pixels to be exposed, resulting in a break in the imaged pattern. For example, if the desired pattern is a circuit, gaps may be inadvertently exposed and formed in a trace, resulting in an electrical gap. The exposure gaps caused by the pixel nature of the micromirror arrays, or pixelation, may cause open circuits or unwanted capacitive effects where trace width or thickness is critical.
Another problem with current art systems is the phenomenon of “stiction,” wherein the individual mirrors in a micromirror area tend to “stick” in a specific orientation if left in that position for an extended period. Consequently, a higher voltage needs to be applied to the mirror drive to point the mirror in another desired direction. Thus the micromirror array consumes more power than normal and affects the reliability of the mirror.
It is known in the art to use gray scale masks in photolithography to form continuously variable material profiles on substrates, such as microlens arrays (wherein each lens can have a different profile), refractive and diffractive micro-optics, precision tapered structures, sinusoidal transmittance gratings, arbitrary shaped micro-optics, and other 3D microstructures, including optical micro-electromechanical devices (MEMS). One type of gray scale mask is a halftone chrome mask that consists, for example, of mixtures of 0.5 micron×0.5 micron chrome spots which are totally opaque and 0.5 micron×0.5 micron clear spots which are totally transparent due to the absence of chrome film coating on the glass photomask substrate. The transmittance of a gray scale resolution element in a halftone chrome mask is determined by the ratio of the number of chrome spots to clear spots. Transmittance decreases as the ratio of chrome spots to clear spots is increased. For a gray scale chrome mask capable of 16 gray levels, a gray scale resolution element must consist of 16 binary spots. A binary spot of the chrome mask is either a chrome spot or a clear spot. When all 16 binary spots in a gray scale resolution element are chrome spots, the gray scale resolution element is totally opaque. However, the images produced using this method are halftone images, not true gray scale images.
Gray scale masks have also been implemented in a photo-emulsion film or a photographic emulsion glass plate and are halftone gray scale patterns, since each silver grain in a developed photographic emulsion is totally opaque. The gray scale is produced by varying the number density of the silver grains. The spacing between the grains are transparent. The size of silver grains is not uniform and may range from, for example, about 0.1 micron to about 1 micron in a high-resolution photoemulsion plate. Therefore, it is difficult to obtain consistent imaging results because of the nonuniformity of grain size.
Yet another type of gray scale is the High Energy Beam Sensitive (HEBS) glass, manufactured, for example by Canyon Material, Inc. In the HEBS glass, process glass photomasks having varying transmissive properties are created in photosensitive glass according to the energy density of a high energy beam impinging on the glass surface. The resulting varying transmissibility glass substrate is used as a gray scale photomask in standard photolithographic processes to create micro-optical elements such as refractive micro lens arrays, diffractive optical elements, prism couples, and three-dimensional micro structures.
Although the HEBS glass process allows true gray scale imaging, a photomask must still be used for photolithographic processing of substrates. While effective, the use of physical masks in photolithography has numerous drawbacks, including the cost of fabricating masks, the time required to produce the sets of masks needed to fabricate semiconductors, the diffraction effects resulting from light from a light source being diffracted from opaque portions of the mask, registration errors during mask alignment for multilevel patterns, color centers formed in the mask substrate, defects in the mask, the necessity for periodic cleaning and the deterioration of the mask as a consequence of continuous cleaning.
It also known in the art to use maskless gray scale x-ray lithography. Maskless gray-scale x-ray lithography has been disclosed (Frank Hartley, Maskless Gray-Scale X-ray Lithography, NASA Tech Brief # NPO-20445, July, 2000). In this reference, a photoresist coated substrate to be patterned is exposed to a parallel beam of hard x-rays. The photoresist is translated across the beam at a varying rate to effect one-dimensional spatial variations in the radiation dose received by the photoresist. The radiation dose delivered to the photoresist on a substrate is made to vary spatially, within a range in which the solubility of the exposed photoresist in a developer liquid varies with the dose.
In conventional gray-scale x-ray lithography, the required spatial variation in the dose is achieved by use of a mask. The mask and the photoresist-covered substrate are translated as a unit across an x-ray beam at a constant rate to obtain the required integrated dose to the mask. In the disclosed maskless technique, the photoresist is not masked. The gradients in the radiation dose needed to obtain gradients in the density of the developed photoresist are generated by controlled variations in the rate of translation of the x-ray beam across the photoresist. These controlled variations define the desired features (variations of the height of the subsequently developed photoresist) to within sub-micron dimensions, depending on the exposure time of the substrate.
After exposure to x-rays, the photoresist coated substrate is developed in the customary manner. After development, the photoresist is dried, giving rise to spatial consolidation of the photoresist into thickness gradients corresponding to the density gradients. The dosage gradients are chosen to achieve desired final thickness gradients, for example, to produce triangular- or saw tooth-cross-section blazes for diffraction gratings. However, the x-rays are inherently dangerous and their use is highly regulated, requiring sophisticated equipment to generate and direct the radiation. In addition, the disclosed techniques require complex translational stages to expose and generate patterns on substrates.
In addition, it is also known in the art to use lasers to create patterns on large photoresist coated substrates. In particular, as disclosed in R. Bawn, et al., “Micromachining System Accommodates Large Wafers,” Laser Focus World, January 2001, pp. 189–192, and “Laser Microfabrication Process,” Proceedings of ICALEO 2000, Paper A49, October 2000, laser lithography techniques can be used to create patterns on large substrates. In the disclosed systems, laser-micromachining systems for large area patterning require a laser source, optics for conditioning and focusing the beam, and a way to precisely control and point the beam. Patterns are produced on substrates by precisely positioning and focusing the laser beam over small area and ablating away the substrate material to form a desired pattern. The laser is then repositioned and another area of the substrate is ablated. The process is continued until the desired pattern on a substrate has been created. In this manner a large substrate can be sequential processed to create large patterned substrates. However, only small areas can be written at one time because of the small beamwidth of the laser, making large area patterning prohibitively time consuming.
Accordingly, there is a need in the art for a method and system for maskless photolithography to provide a more effective way to fabricate custom devices in a low volume production environment. This system needs to combine ease of use, reconfigurability, and the ability to provide coarse manual alignment and automated fine alignment of mask patterns. In addition the system needs to address the exposure gaps inherent in the process due to the pixel nature of the projected mask and provide means for eliminating stiction. In summary, the system needs to provide all the advantages of a maskless photolithography system at a reasonable cost, and include capabilities tailored to direct writing in a rapid prototyping environment.
In addition, there is a need in the art for a method and system for maskless photolithography to provide gray scale capability and large area patterning of substrates. This system can combine ease of use, reconfigurability, and the ability to provide gray scale patterns to create variable thickness on exposed substrates. In addition the system can allow patterning large areas in a single exposure. The system can provide many of the advantages of a maskless photolithography system at a reasonable cost, and include capabilities tailored to gray scale imaging and large area pattern generation in a cost effective, easy to implement system.